Ai predicts mental health crises from sparse digital data
Peer-Reviewed Publication
Updates every hour. Last Updated: 13-Aug-2025 16:11 ET (13-Aug-2025 20:11 GMT/UTC)
Researchers present a machine learning framework that forecasts individual mental health deterioration using limited, real-world data from wearables and smartphones. This enables personalized early interventions, shifting psychiatry from reactive to proactive care.
A group of 15 leading scientists from academia and industry published a paper that provides practical guidance on postbiotic quantification that can be implemented worldwide. Rather than presenting a rigid set of rules, the publication provides a decision tree to assist in the selection of the most appropriate quantification techniques—depending on the type of microorganisms, cell integrity, and number of strains present—to ensure that postbiotic products are accurately characterized and can be reproducibly manufactured.
Researchers have successfully demonstrated that advanced generative AI (GenAI) models can accurately assess lung adenocarcinoma pathological features with remarkable precision. The comprehensive study shows Claude-3.5-Sonnet achieving 82.3% accuracy in cancer grading, potentially revolutionizing how pathologists diagnose and predict outcomes for lung cancer patients.
A new study in Nature that finds a gene therapy against HIV given at birth can protect breastfeeding babies from infection with a single dose.